Adaptive Filter Algorithms Based Noise Cancellation Using Neural Network in Mobile Applications

Noise-free output is a desired characteristic of any mobile communication system. Adaptive noise cancellation is achieved by subtracting unwanted noise signal from the corrupted signal. We propose signal extraction using artificial neural network hybrid back propagation adaptive for mobile systems. The performance analysis of the proposed hybrid adaptive algorithms is carried out based on the error convergence and correlation coefficient. By taking into consideration of the existing algorithms, the proposed algorithms require small neural training sets and it gives good results. Noise cancellation operation is established through adaptive control with the goal of achieving minimum noise error level at output. This paper focuses on the analysis of noise cancellation using least mean square algorithms, gradient adaptive lattice algorithms, and hybrid adaptive algorithms. From computed output, we observed that the hybrid adaptive algorithms perform better.